In the annual Doing Business (DB) report, the World Bank publishes indicators that quantify the regulatory barriers that are present on 11 areas of the business ‘life cycle’, such as the enforcement of contracts, the payment of taxes and the resolution of insolvencies. These indicators allow policymakers to compare business regulatory environments across time and countries, and to identify unnecessarily lengthy or costly administrative procedures.

The data from this report is widely used. Both the World Bank and the IMF refer to the DB indicators in their reports and more than 2000 research papers that used the DB indicators had been published in academic journals by 2013 (Besley, 2015). It is estimated that the website of the report received close to 3 million visits in 2012 (World Bank, 2012). As a result of this, the report exerts significant influence in the public policy debate. More than 60 countries like China and India have government units devoted to improving their positions in the DB ranking (“Doing BRICness”, 2018); and, according to the World Bank, “Governments have reported more than 3,500 regulatory reforms, 1,116 of which have been informed by Doing Business since 2003” (World Bank, 2019a, pg. 29).

In some cases, it seems as if climbing up in the DB ranking was deemed as a valuable end in itself. While this may sometimes spark socially desirable reforms, such a narrow focus on these metrics might lead policymakers astray in other cases. In this article, we argue that because the Total tax and Contribution rate is not a comprehensive measure of the tax system, reforms that lead towards ‘better’ scores in the indicator are not always socially desirable. We also point to methodological problems of this indicator that limit the accuracy with which the Doing Business report measures the costs induced by tax systems.

Paying Taxes

One of the eleven main indicators published in the DB report is Paying Taxes, which measures the administrative and tax burdens that tax systems impose on medium-sized firms. The score that each country receives in this indicator is the unweighted average of the scores that it receives in each of the four Paying Taxes component indicators, and the scores in these component indicators are in turn computed using data about the situation faced by a standardized case study firm that performs industrial or commercial activities, does not participate in foreign trade, operates in the largest business city of the country, has 60 employees and a turnover of 1050 times income per capita. [1]

The first of the component indicators, Time, measures in hours the time that it takes the fictitious firm to prepare, file and pay the corporate income tax, the VAT or sales tax, and labor taxes. The second component indicator, Postfiling index, measures in weeks the time that it takes to request and process a VAT refund claim, as well as the time that it takes to comply with and complete a corporate income tax correction. Tax payments, the third component indicator, “reflects the total number of taxes and contributions paid, the method of payment, the frequency of payment, the frequency of filing and the number of agencies involved for the standardized case study company during the second year of operation.” (World Bank, 2019a).

The fourth component indicator, Total tax and Contribution rate, intends to measure the tax burden that falls on the case study firm during its second year of operation, and it is expressed as a percentage of pre-tax profits. In order to compute this indicator, both the total amount of taxes and contributions borne and the pre-tax benefits are needed. The former amount is equal to “the sum of all the different taxes and contributions payable after accounting for allowable deductions and exemptions” (World Bank, 2019a, pg.105). Due to the fact that those taxes that are not legally borne by the firms are excluded from the calculation, taxes such as the personal income tax, the VAT tax and the sales tax are not considered by this indicator. To calculate the corporate income tax amount, pre-tax profits are defined as “sales minus cost of goods sold, minus gross salaries, minus administrative expenses, minus other expenses, minus provisions, plus capital gains (from the property sale) minus interest expense, plus interest income and minus commercial depreciation”, but when it comes to dividing the total amount of taxes by pre-tax profits, the latter are always assumed to be equal to 59.4 times income per capita (World Bank, 2019a).

Because each one of these sub-indicators indicators is expressed in a different unit of measurement, they are all normalized to a common unit in order to calculate the Paying Taxes indicator (which is then used along the other main indicators to compute the overall Doing Business score). In the case of the Total tax and Contribution rate, the ‘normalized’ score is determined using the following formula:

Source: Morck & Shou (2018), pg 20.

Where t is the tax burden of the country whose score is being calculated, t_L is the ‘best regulatory performance’, which is equal to the highest total tax and contribution rate among the 15% of economies with the lowest total tax and contribution rate in the Doing Business sample for all years included in the analysis up to and including Doing Business 2015 (26% of pre-tax benefits), and t_H  (the ‘worst regulatory performance’) is the 95th percentile among all economies in the Doing Business sample (84%).

Total tax and Contribution rate indicator: what does it measure?

In the public debate, policy reforms that lead to higher positions in the DB ranking tend to be framed in a positive way, whereas drops in the ranking are interpreted as signs of negative developments. Although all reforms entail both benefits and costs, this interpretive framework will hardly lead to wrong conclusions if changes in the DB ranking stem from reforms that are guided by the objective of improving the score in some DB indicators such as the one that measures the time that it takes to start a new business or the one that captures the median time that is necessary to obtain an electricity connection.

This interpretative framework is much less likely to lead to correct assessments when changes in the DB ranking are rooted in changes in the Total tax and Contribution rate indicator. The theory of optimal taxation, a tool that is frequently used to design and assess the consequences of tax reforms, posits that tax systems should be designed to maximize a social welfare function that aggregates the individual preferences of the citizens (Mankiw, Weinzierl & Yagan, 2009). While there is a large variety of social welfare functions that might be used in these exercises, researchers usually choose functions that - holding fixed the total sum of individual utilities - rank more egalitarian distributions as better than less egalitarian distributions, acknowledging in this way that while efficiency matters, there is a social preference for equity. The methodology of the Total tax and Contribution rate indicator is oblivious to this literature. Indeed, the Bank warns that “[...] the indicators do not measure—nor are they intended to measure—the benefits of the social and economic programs funded with tax revenues.” (World Bank, 2019a, pg.24).

Although this means that judging whether tax reforms are socially desirable or not by analyzing changes in the Total tax and Contribution rate score is a recipe for mistakes, it could be argued that the purpose of this indicator is less ambitious. On the ‘Why tax rates matter’ section of the DB website, the Bank states that tax rates matter because higher corporate tax burdens are associated with fewer formal businesses and lower private investment (World Bank, 2019b). Because of this, the Total tax and Contribution rate – originally developed by PwC - might be better thought of as a measure of the legal tax burden that was included in the report with the ultimate goal of reflecting the distortions generated by tax systems. As such, this indicator should be judged by whether it measures the tax burden properly and by whether it helps or not to assess if a tax system gets more or less distortive after a tax reform.

Regarding the first of these two points, we argue that the Total tax and Contribution rate is a flawed measure of the legal tax burden. The reason is simple: while this indicator is supposed to measure firm tax liability expressed as a percentage of pre-tax benefits, its methodology fails to acknowledge the fact that both tax bases and pre-tax benefits change in response to changes in the tax system.

Example: tax burden of a payroll tax

Suppose a firm has an annual turnover of 1050 times income per capita (henceforth $) and only contracts labor. Since the firm hires 60 workers and pays an average monthly wage of 1.2$, its annual costs equal 864$ and it has annual pre-tax benefits of 186$ (1050$-864$). [2]

In a tax-less world, the tax burden that falls on this firm would undoubtedly be 0%. What would the tax burden be if a payroll tax of 23% were levied on gross wages?

In the World Bank exercise, tax bases and pre-tax benefits are assumed to be invariant to the introduction of taxes, so the gross wages and pre-tax benefits that are used to answer this question still amount to 864$ and 186$, respectively. In consequence, the standardized case study firm pays 198.7$ (864$*23%*100) in concept of payroll taxes, and the tax burden that is yielded by the Doing Business methodology is equal to 106.8% of pre-tax benefits (198.7$/186$).

However, the introduction of taxes tends to produce changes in tax bases and pre-tax benefits, as taxes modify equilibrium prices and quantities. The payroll tax introduces a wedge between the cost of labor of a firm and gross salaries which can be represented as a downward shift in the labor demand schedule, as the gross salaries that firms are willing to pay to hire a certain amount of labor decrease. Standard partial equilibrium incidence analysis states that regardless of who has to write the check to the tax authority, the extent to which the cost of labor increases and gross salaries decrease depends on the elasticities of labor demand and supply. For example, if labor supply is perfectly inelastic, the introduction of a payroll tax leads to lower gross salaries, while the total cost of labor and the amount of labor hired by the firm do not change.

Coming back to our example, the introduction of a payroll tax rate of 23% on gross salaries leads to a reduction of 18,7% in the gross salaries that the firm is willing to pay to hire a certain amount of labor. If we assume that the labor supply is perfectly inelastic, the payroll tax reduces gross wages by 18,7% as well. While the total cost of labor remains unchanged, gross salaries decrease to 702.4$, and pre-tax profits increase to 347.6$ (1050$ - 702.4$). Taking into account that the firm pays an amount equal to 161.6$ in concept of payroll taxes (23%*702.4$), the tax burden induced by this tax is equivalent to 46,5% of pre-tax benefits, a considerably lower figure that the one yielded by the DB methodology (106.8% of pre-tax benefits), which bias upward the legal tax burden by 129%.

Furthermore, even when it is computed considering the economic consequences of taxes, the legal tax burden overestimates the economic incidence of taxes on firms. In our example, the entirety of the payroll tax is shifted to workers, who receive lower wages: neither the behavior nor the after-tax profits of the firm are different from what they would be in a tax-less world.

Total tax and Contribution rate as an efficiency proxy

Although the Total tax and Contribution rate constitutes a flawed measure of the legal tax burden, it provides useful information to assess whether the efficiency of a tax system has decreased or augmented after a tax reform. Holding everything else fixed, systems based on VAT taxes get higher scores than those based on highly distortive turnover taxes. In addition to this, its score rises with the lowering of corporate taxes and it heavily penalizes the introduction of taxes on bank deposits and withdrawals. Despite this, the way in which this indicator is normalized compromises the accuracy with which the overall Doing Business indicator (which is used to generate the DB ranking) reflects changes to the tax system.

Due to the fact that the DB methodology does not take into account that tax bases and pre-tax benefits are not invariant to taxes, legal tax burdens calculated using this methodology sometimes adopt extreme values. Between 2010 and 2012, for example, according to the World Bank, the tax burden imposed on firms operating in the Democratic Republic of Congo was equal to 339.1% of pre-tax benefits. If outliers like this are not taken care of when rescaling the data to compute the normalized scores, the distribution of the Total tax and Contribution rate scores ends up being tilted towards high values. To avoid this, the World Bank sets the ‘worst performance’ at the 95th percentile of the tax burden distribution (84%), and all the economies that have a tax burden over 84% get a score equal to 0. As a side effect of this procedure, changes in the scores of the overall DB indicator underestimate the effects of those reforms that are implemented in countries where the initial or the final legal tax burden is over 84%.

A similar situation arises at the opposite tail of the tax burden distribution. With the objective of reducing the bias towards countries that do not rely on taxes on firms similar to the standardized case study firm, all countries with tax burdens lower than 26.1% obtain the highest score in the Total tax and Contribution rate indicator (100). In consequence, the effects of tax reforms implemented in countries that start or end up with tax burdens below this threshold are not fully reflected by movements in the score of the overall Doing Business indicator. Sometimes, they are not reflected at all.

Conclusion

In this article we have explained why it is that, as was indicated in the 2008 Independent Evaluation of the Doing Business Indicators, ‘better’ scores in the Total tax and Contribution rate indicator do not necessarily reflect increases in social welfare. As this indicator is intended to measure the costs that tax systems impose on private firms but it is oblivious to distributional matters and does not capture the benefits associated with higher tax revenues, rigorous assessments of tax reforms cannot be based on it.

In addition to this, we have pointed to methodological problems of the Total tax and Contribution rate indicator that undermine its usefulness as a simple measurement of the tax burden on firms. These issues end up compromising the entire Doing Business report. Firstly, we have shown that owing to the fact that the methodology omits that both tax bases and pre-tax profits change with the introduction of taxes, the tax burden that is calculated by the World Bank does not correctly reflect the legal tax burden that actually falls on medium-sized enterprises. Besides this, we have shown that the legal tax burden associated to a certain tax differs from its economic burden, which is the one relevant to understand the behavior of firms. Secondly, we pointed out that while the Total tax and Contribution rate can be reasonably expected to reflect whether tax reforms lead to less distortive tax systems, the fact that countries with tax burdens of more than 84% of pre-tax benefits get the lowest score in this indicator and countries with tax burdens lower than 26.1% get the maximum score implies that the effects of tax reforms are not always fully reflected by changes in the overall Doing Business indicator.

1. A more comprehensive list of assumptions is provided in the following website: https://www.doingbusiness.org/en/methodology/paying-taxes

2. In our example, we depart from the World Bank assumption that pre-tax benefits equal 59.4 times income per capita. This does not change the conclusion of the analysis, which depends only on the assumption that pre-tax benefits do not change in response to tax changes.

References:

Besley, T. (2015). Law, Regulation, and the Business Climate: The Nature and Influence of the World Bank Doing Business Project. Journal of Economic Perspectives, 29(3), 99–120.

Doing BRICness (2018, November 3). The Economist.

Mankiw, N. G., Weinzierl, M., & Yagan, D. (2009). Optimal Taxation in Theory and Practice. Journal of Economic Perspectives, 23(4), 147–174.

Morck, R. & Shou, J.C (2018). On the Integrity of the “Ease of Doing Business” Indicators.

World Bank. (2012). Executive Board Informal Meeting: Doing Business update. 12 July, p. 1‐34.

World Bank. (2019a). Doing Business 2019: Training for Reform. Washington, DC: World Bank. DOI: 10.1596/978-1-4648-1326-9. License: Creative Commons Attribution CC BY 3.0 IGO

World Bank. (2019b). Paying Taxes: why it matters. https://www.doingbusiness.org/en/data/exploretopics/paying-taxes/why-matters